Creating Pivots based for a rowcount in SQL - sql

I have 2 tables T1 and T2.
T1 has the following columns:
Id, EntityId, TypeofValue, Year, value
EntityId can have 7 values:
1,2,3,4,5,100,101
Typeofvalue can have 2 values
1 indicates Actual
2 indicates Target
T2 has the following columns:
NoOfRecordsToDisplay
I need to fetch the number of records (if existing) for Target corresponding to an Id. The record count (1 if any record is present 0 if none) for Target needs to be grouped into two categories: First having Entityid 1 and Maximum records under Second group with entityids 2,3,4,100,101 [Not 5]
However, the catches are:
Sometimes Target value might not be present for a year
I need to get only last records for targets on the basis of NoOfRecordsToDisplay (The number of records to display comes from T2) for actual
Example1:
NoOfRecordsToDisplay =3, ID =123
The data below should return
CountGroup1: 1, Countgroup2: 1
as Entityid 1 has least one value for target for last 3 years -2015, 2014,2013 in this case
as Entityid 2 or 3 has at least 1 value for years -2015, 2014
Id EntityId TypeofValue Year Value
123 1 1 2015 55
123 1 1 2014 56
123 1 1 2013 57
123 1 1 2012 58
123 1 2 2015 50
123 1 2 2014 50
123 1 2 2013 50
123 1 2 2012 50
123 2 1 2015 55
123 2 1 2014 56
123 3 1 2015 57
123 3 1 2014 58
123 2 2 2015 55
123 2 2 2014 56
123 3 2 2015 57
124 1 1 2015 55
124 1 1 2014 56
124 2 1 2013 57
124 2 1 2012 58
124 1 2 2015 50
124 1 2 2014 50
124 2 2 2013 50
124 2 2 2012 50
Another dataset
NoOfRecordsToDisplay =3, ID =123
The data below should return:
CountGroup1: 0, Countgroup2: 1
as Entityid 1 has no target value for last 3 years (entityid 1 has a target value but for 2012)
as Entityid 2 has at one value for target years -2015 (entityid 3 has a target value but for 2010)
Id EntityId TypeofValue Year Value
123 1 1 2015 55
123 1 1 2014 56
123 1 1 2013 57
123 1 1 2012 58
123 1 2 2012 58
123 2 1 2015 55
123 2 1 2014 56
123 2 2 2015 55
123 2 2 2011 56
123 3 2 2010 57
Thank you so much for your help.
I have been trying to find this solution for a long time, I am not sure if Pivot will help
The question is different from the other one that I had posted as I am trying to create a group count based on entity groups.

Related

sql split yearly record into 12 monthly records

I am trying to use common table expression to split an yearly record into 12 monthly records. I have to do it for next 20 years records . That means 20 rows into 600 rows (20*12=600 records).
What is the best way to do it. Can anyone help with an efficient way to do it.
Using a single table as shown below. Year 0 means current year so it should split into remaining months and year=1 means next year onward it should split into 12 (months) records
id year value
1 0 3155174.87
1 1 30423037.3
1 2 35339631.25
expected result should look like this:
Id Year Month Value Calender year
1 0 5 150 2022
1 0 6 150 2022
1 0 7 150 2022
1 0 8 150 2022
1 0 9 150 2022
1 0 10 150 2022
1 0 11 150 2022
1 0 12 150 2022
1 0 1 150 2023
1 0 2 150 2023
1 0 3 150 2023
1 0 4 150 2023
1 1 5 100 2023
1 1 6 100 2023
1 1 7 100 2023
1 1 8 100 2023
1 1 9 100 2023
1 1 10 100 2023
1 1 11 100 2023
1 1 12 100 2023
1 1 1 100 2024
1 1 2 100 2024
1 1 3 100 2024
1 1 4 100 2024

You can simply join onto a list of months, and then use a bit of arithmetic to split the Value
SELECT
t.Id,
t.Year,
v.Month,
Value = t.Value / CASE WHEN t.Year = 0 THEN 13 - MONTH(GETDATE()) ELSE 12 END
FROM YourTable t
JOIN (VALUES
(1),(2),(3),(4),(5),(6),(7),(8),(9),(10),(11),(12)
) v(Month) ON t.year > 0 OR v.Month >= MONTH(GETDATE());
db<>fiddle

Transposing multiple related columns

While transposing single columns is pretty straight forward I need to transpose a large amount of data with 3 sets of , 10+ related columns needed to be transposed.
create table test
(month int,year int,po1 int,po2 int,ro1 int,ro2 int,mo1 int,mo2 int, mo3 int);
insert into test
values
(5,2013,100,20,10,1,3,4,5),(4,2014,200,30,20,2,4,5,6),(6,2015,200,80,30,3,5,6,7) ;
select * FROM test;
gives
month
year
po1
po2
ro1
ro2
mo1
mo2
mo3
5
2013
100
20
10
1
3
4
5
4
2014
200
30
20
2
4
5
6
6
2015
200
80
30
3
5
6
7
Transposing using UNPIVOT
select
month, year,
PO, RO, MO
from ( SELECT * from test) src
unpivot
( PO for Description in (po1, po2))unpiv1
unpivot
(RO for Description1 in (ro1, ro2)) unpiv2
unpivot
(MO for Description2 in (mo1, mo2, mo3)) unpiv3
order by year
Gives me this
month
year
PO
RO
MO
5
2013
100
10
3
5
2013
100
10
4
5
2013
100
10
5
5
2013
100
1
3
5
2013
100
1
4
5
2013
100
1
5
5
2013
20
10
3
5
2013
20
10
4
5
2013
20
10
5
5
2013
20
1
3
5
2013
20
1
4
5
2013
20
1
5
4
2014
200
20
4
4
2014
200
20
5
4
2014
200
20
6
4
2014
200
2
4
4
2014
200
2
5
4
2014
200
2
6
4
2014
30
20
4
4
2014
30
20
5
4
2014
30
20
6
4
2014
30
2
4
4
2014
30
2
5
4
2014
30
2
6
6
2015
200
30
5
6
2015
200
30
6
6
2015
200
30
7
6
2015
200
3
5
6
2015
200
3
6
6
2015
200
3
7
6
2015
80
30
5
6
2015
80
30
6
6
2015
80
30
7
6
2015
80
3
5
6
2015
80
3
6
6
2015
80
3
7
I will like to turn it to something like this. Is that possible?
month
year
PO
RO
MO
5
2013
100
10
3
5
2013
20
1
4
5
2013
0
0
5
4
2014
200
20
4
4
2014
30
2
5
4
2014
0
0
6
6
2015
200
30
5
6
2015
80
3
6
6
2015
0
0
7

Maybe use a query like below which creates rows as per your design using CROSS APPLY
select month,year,po,ro,mo from
test cross apply
(values (po1,ro1,mo1), (po2,ro2,mo2),(0,0,mo3))v(po,ro,mo)
see demo here

Unpivot acts similar as union,Use union all in your case
SELECT month,
year,
po1 AS PO,
ro1 AS RO,
mo1 AS MO
FROM test
UNION ALL
SELECT month,
year,
po2,
ro2,
mo2
FROM test
UNION ALL
SELECT month,
year,
0,
0,
mo2
FROM test

Want to retreive data in required format from SQL SERVER table

I have a view as follows:
id paperid papercode papername marks1 marks2 total
1 1 001 paper-1 10 35 45
2 2 001 paper-2 12 40 52
3 3 002 paper-3 9 30 39
Now, I want data as below:
id paperid papercode papername marks1 marks2 total
1 1 001 001 22 75 97
3 3 002 paper-3 9 30 39

You can try below -
select min(id),min(paperid),papercode,min(papername),sum(marks1),sum(marks2),sum(marks3)
from tablename
group by papercode

SUM column values based on two rows in the same tables in SQL

I have one table like below in my SQL server.
Trans_id br_code bill_no amount
1 22 111 10
2 22 111 20
3 22 111 30
4 22 111 40
5 22 111 10
6 23 112 20
7 23 112 20
8 23 112 20
9 23 112 30
and I want desired output like below table
s.no br_code bill_no amount
1 22 111 110
2 23 112 90

try this:
select br_code, bill_no, sum(amount)
from TABLE
group by br_code, bill_no

SQL terminology to combine a NOT EXIST query with latest value

I am a beginner with basic knowledge.
I have a single table that I am trying to pull all UID's that have not had a particular code in the table within the past year.
My table looks like this: (but much larger of course)
FACID DPID EID DID UID DT Code Units Charge ET Ord
1 1 6 2 1002 15-Mar-07 99204 1 180 09:36.7 1
1 1 7 5 10004 15-Mar-07 99213 1 68 02:36.9 1
1 1 24 55 25887 15-Mar-07 99213 1 68 43:55.3 1
1 1 25 2 355688 15-Mar-07 99213 1 68 53:20.2 1
1 1 26 5 555654 15-Mar-07 99213 1 68 42:22.6 1
1 1 27 44 135514 15-Mar-07 99213 1 68 00:36.8 1
1 1 28 2 3244522 15-Mar-07 99214 1 98 34:59.4 1
1 1 29 5 235445 15-Mar-07 99213 1 68 56:42.1 1
1 1 30 3 3214444 15-Mar-07 99213 1 68 54:56.5 1
1 1 33 1 221444 15-Mar-07 99204 1 180 37:44.5 1
I am attempting to use the following, but this is not working for my time frame limits.
select distinct UID from PtProcTbl
where DT<'20120101'
and NOT EXISTS (Select Distinct UID
where Code in ('99203','99204','99205','99213',
'99214','99215','99244','99245'))
I need to know how to make sure the UID's that I am pulling are the ones don't have a DT after the 1/1/2012 cut off date that contains one of the NOT Exists codes.
The above query returned UID's that actually dates after 1/1/2012 that does contain one of the above codes...
Not sure what I am doing wrong or if I am totally off base on this..
Thanks in advance.

Are you sure you need the NOT EXISTS? How about instead:
AND Code NOT IN ('99203','99204','99205','99213','99214','99215','99244','99245')